This paper presents a new incremental learning solution for Linear Discriminant Analysis (LDA). We apply the concept of the sufficient spanning set approximation in each update st...
Recently, boosting is used widely in object detection applications because of its impressive performance in both speed and accuracy. However, learning weak classifiers which is on...
Similarity metrics that are learned from labeled training
data can be advantageous in terms of performance
and/or efficiency. These learned metrics can then be used
in conjuncti...
Chord sequences are a compact and useful description of music, representing each beat or measure in terms of a likely distribution over individual notes without specifying the not...
Recently, boosting has come to be used widely in object-detection applications because of its impressive performance in both speed and accuracy. However, learning weak classifier...